Fire-resistant glazing

Fire-resistant glazing is glazing that, together with its frame, fixings and glazing beads, has a specified fire resistance rating. It can therefore form part of a fire separation or a fire-rated external wall. New Zealand's Verification Method C/VM2 (second edition, 2025) calls it fire resisting glazing and defines it as a complete fixed or openable assembly, not as a type of glass alone. Its main difference from ordinary windows is that it stays in place during a fire for its rated time. Ordinary exterior windows are assumed in C/VM2 to fall out at an upper layer temperature of 500 °C or when the fire becomes limited by ventilation.

Fire-resistant glazing lets designers keep light and views in places where a solid wall would otherwise be needed. Typical uses in the sources are vision panels in fire doors, windows in external walls close to a boundary, and glazed panels beside escape routes. Some products only keep flames and hot gases back, while insulated products also limit the heat on the unexposed side. The difference matters for people who pass close to the glass, because heat radiation through uninsulated glazing can still be high.

Ordinary glass, uninsulated fire-resistant glazing and insulated fire-resistant glazing in a fully developed fire. Based on Ministry of Business, Innovation and Employment (2025b), pp. 16 and 35.

Fire resistance criteria for glazing

Glazing is rated by the same fire resistance criteria as walls and doors. In New Zealand, a fire resistance rating (FRR) gives three times in minutes, for structural adequacy, integrity and insulation, always in that order. Integrity is the time for which a test specimen prevents the passage of flame or hot gases. Insulation is the time for which it limits the transmission of heat through the specimen. These times come from a furnace test that follows the standard fire curve. The Introduction to Fire Safety Engineering stresses that this curve is an agreed temperature course for comparing constructions, not a model of a real fire.

The Australian data sheets give the criteria of the test standard AS 1530.4 in some detail. Integrity is lost when a cotton pad held over a crack ignites, or when a through gap larger than 150 mm by 6 mm or 25 mm in diameter appears. Flaming for 10 s or more on the unexposed face also counts as failure. Insulation is lost when the average temperature of the unexposed face rises by 140 K, or the maximum temperature at any point by 180 K. A fire door with glazing must also not fail by radiation through any glazed part during its integrity period.

For glazed areas in fire doors, AS 1530.4 and the Australian National Construction Code add a separate radiation criterion. The heat flux at 365 mm from the vision panel may not exceed 10 kW/m² for the period of the integrity rating. At that level, combustibles stored more than 365 mm from the door are not expected to ignite, and the measured flux also shows the risk to people walking past. The limit makes the role of thermal radiation explicit: integrity alone does not stop heat from passing through clear glass.

In the United States, NUREG-1805 summarises the window test standard NFPA 257. Fire ratings of windows were once limited by the failure of wired glass at about 870 °C (1,600 °F). Advances in glazing technology have since raised the available ratings of window assemblies to 20, 30 and 45 minutes. Ratings of 60, 90 and 180 minutes or more are available under the general fire resistance test NFPA 251. Within 2 minutes after the fire test, the fire-exposed side of the window assembly is subjected to a standard hose stream.

Behaviour compared with ordinary glass

Ordinary window glass usually breaks in a fully developed fire, and its failure changes the fire. C/VM2 (second edition, 2025) assumes that exterior windows without fire resistance break and fall out at the earlier of two moments. These are an average upper layer temperature of 500 °C and the point at which the fire becomes ventilation-limited. A comment in C/VM2 explains that it is hard to predict when glass falls out, and that 500 °C represents a rapid onset of flashover. Windows with non-fire-resisting glazing are therefore counted as openings that supply air to the fire when the fire severity is calculated. Fire-resistant glazing, in contrast, may be assumed to stay in place up to its rated time.

The International Fire Engineering Guidelines make a similar distinction. For analysis, fire may be assumed to spread through a closed window at flashover if the window is of ordinary glass, but not if it is toughened or wired. They add that fire can spread to an adjacent building through the glass of a fixed or closed window by radiation, without any breakage. More on how and when glass fails is given in Glass breakage in fire. In zone models, the failure of glass or polycarbonate parts of the building envelope can be included as the opening of a vent at a set temperature or rate of temperature rise.

A NIST test for the World Trade Center investigation shows the difference between glass types. A 2.4 m by 2.4 m window assembly held two double-pane tempered glass units with a gel insulation layer between the panes and two single-pane tempered units. The lower single-pane unit cracked and fell out about 185 s after ignition and the upper one about 212 s after ignition. The double-pane units never fell out. The extra ventilation from the broken panes probably caused the highest peak heat release rate of the test series, almost 16 MW.

Glazing also matters for smoke. In C/VM2, smoke separations that include glazing and comply with the relevant product specification stay in place up to their rated temperature or until flashover, whichever comes first. A separation that consists simply of toughened glass, not designed to a standard, is assumed to disappear once the average upper layer temperature reaches 200 °C.

Radiation through glazing along escape routes

People who escape past a window or glazed panel must not receive so much radiation that they feel pain. C/VM2 (second edition, 2025) sets a maximum exposure of 10 kW/m² for occupants in exitways or on external escape routes. Without fire-resistant glazing, all occupants must also have passed the window within 10 minutes after ignition. The method calculates the radiation at the point opposite the centre of the window, at a height of 2.0 m or at mid-height of the glazing, whichever is lower. It multiplies the heat flux emitted by the window by a view factor, the emissivity of the fire gases and a glazing factor:

$${\dot{q}}^{\prime }{\text{'}}_r=F_w\epsilon k{\dot{q}}^{\prime }{\text{'}}_w$$

Symbol	Description	Unit
${\dot{q}}^{\prime }{\text{'}}_r$	maximum incident radiation at the escape route	kW/m²
$F_w$	view factor from the window to the point on the escape route	–
$\epsilon$	emissivity of the fire gases, taken as 1.0	–
$k$	glazing factor: 0.5 for fire resisting glazing, 1.0 for all other glazing	–
${\dot{q}}^{\prime }{\text{'}}_w$	design heat flux emitted from the window: 83, 103 or 144 kW/m² for a fire load energy density up to 400, 400 to 800, or above 800 MJ/m²	kW/m²

The glazing factor expresses that radiation through uninsulated fire-resistant glazing may be taken as half that through other glazing. The relation applies to a single window; several windows need a more detailed, time-dependent analysis. It is not meant for occupants who are likely to be mobility-impaired. No analysis is needed where insulated glazing with an FRR of at least -/30/30 is used, or where window wetting sprinklers listed for that purpose are installed.

A worked example takes a glazed panel 2.0 m wide and 2.0 m high in a firecell with a fire load energy density of at most 400 MJ/m², so that ${\dot{q}}^{\prime }{\text{'}}_w$ = 83 kW/m². An escape route passes 1.0 m in front of the panel. Using the table method of the Introduction to Fire Safety Engineering, the panel is split into four squares of 1.0 m by 1.0 m, each with a view factor of 0.139. The total view factor is 4 × 0.139 ≈ 0.56, and with ordinary glass ($k$ = 1.0), the radiation is 0.56 × 1.0 × 1.0 × 83 ≈ 46 kW/m². With uninsulated fire-resistant glazing ($k$ = 0.5) it is about 23 kW/m², still more than twice the limit of 10 kW/m².

The example also shows how distance helps. Calculating the view factor for larger distances shows that the radiation from this panel falls to 10 kW/m² at about 2.0 m with uninsulated fire-resistant glazing, and only at about 3.0 m with ordinary glass. At 1.0 m the design needs another solution, such as insulated glazing of at least -/30/30, listed window wetting sprinklers or an alternative escape route.

Radiation opposite the centre of a 2.0 m by 2.0 m glazed panel with an emitted flux of 83 kW/m², for ordinary glass (k = 1.0) and uninsulated fire-resistant glazing (k = 0.5). Based on Ministry of Business, Innovation and Employment (2025b), pp. 34 to 35, and Van Rossum (2027), §5.4.

External walls and boundaries

Glazing in external walls is one of the routes for fire spread between buildings. In New Zealand, building code clause C3.6 limits the radiation received at the relevant boundary to 30 kW/m², and to 16 kW/m² at 1 m beyond it. In the C/VM2 method, non-fire-rated window and door assemblies count as unprotected areas that radiate heat. The design emitted flux from such areas is 83, 103 or 144 kW/m² for unsprinklered firecells, depending on the fire load energy density. For sprinklered firecells the values are lower, at 58, 72 and 101 kW/m². The rest of the wall must keep its integrity through the full burnout design fire and have enough insulation to meet clause C3.7.

The Introduction to Fire Safety Engineering applies the same logic to openings in a facade. Its worked example has two openings of 1.8 m by 1.2 m, 2.4 m apart, that radiate 120 kW/m² towards a facade 6 m away. The radiation received there is 3.84 kW/m², below the limit of 15 kW/m² used in the example, so no further measures are needed. If the result had exceeded the limit, the openings would have to be made fire-resistant or another measure would be needed.

Close to a boundary the rules become strict. Within 1.0 m of a relevant boundary, C/VM2 permits no unprotected areas except small ones of up to 0.1 m² and areas of fire resisting glazing. The glazing must be rated for integrity, and its FRR, like that of the wall, is derived from the full burnout design fire. Neighbouring areas of fire-resistant glazing in the same space must be spaced horizontally by at least the greater of their widths, and vertically by at least the greater of their heights.

The maximum area of each glazed area depends on the distance to the boundary, the fire load and whether the firecell is sprinklered. In an unsprinklered firecell, each area is limited to 1.0 m² up to 0.4 m from the boundary, whatever the fire load. At 1.0 m the limits are 6.0 m², 3.5 m² and 1.5 m² for fire load energy densities up to 400, 400 to 800 and above 800 MJ/m². For a fire load of up to 400 MJ/m², the area is unlimited beyond 1.7 m; for higher fire loads it never exceeds 15 m². Sprinklers relax the limits considerably: with a fire load of up to 400 MJ/m² there is no limit at all.

Maximum area of each fire resisting glazed area in an unsprinklered firecell, by distance to the relevant boundary and fire load energy density (C/VM2, New Zealand). Based on Ministry of Business, Innovation and Employment (2025b), p. 74.

Glazing inside buildings

Inside buildings, fire-resistant glazing forms part of the walls and doors that maintain fire compartmentation. Vision panels in a fire door are one such case, and they are the reason for the radiation criterion of 10 kW/m² at 365 mm in Australian door tests. Glazed panels beside escape routes fall under the C/VM2 radiation check, and any openable part must close automatically when smoke or heat is detected.

Older and current American codes show how glass has long been accepted in protected openings. The BOCA Building Code in force when the World Trade Center was designed allowed openings in shafts to be protected with approved fire doors, curtains, shutters or fixed metal sash with wired glass. The current (2003) New York City Building Code allows atrium enclosure walls of wired, laminated or tempered glass instead of 2 hour rated construction. In that case sprinklers no more than 6 ft (1.8 m) apart must protect the glass on the occupied side.

Limits and uncertainty

A fire resistance rating describes performance in a standard furnace test, not in a real fire. The Introduction to Fire Safety Engineering recommends comparing the thermal load in a natural fire with the load in the fire test. A zone model can estimate when glazing fails if a limit temperature or rate of rise is known. The temperature in a real fire depends on room size, openings, insulation and heat release rate, so a test curve cannot reproduce a specific fire.

Several simplifications in the design rules point to remaining uncertainty. C/VM2 states that it is hard to predict reliably when glass falls out of a window. The fixed glazing factor of 0.5 treats all uninsulated fire-resistant glazing alike, and the escape route method is limited to a single window. The NIST test also showed that the breakage time of ordinary panes can change the ventilation and the heat release rate of a fire.

See also

• Fire curtain
• Intumescent material
• Fire wall
• Passive fire protection

References

• Australian Building Codes Board. (2020a). Data sheets: Fire Safety Verification Method (Handbook annex, Print version 1.1).
• Hamins, A., Maranghides, A., McGrattan, K. B., Ohlemiller, T. J., & Anleitner, R. L. (2005). Experiments and modeling of multiple workstations burning in a compartment (NIST NCSTAR 1-5E). National Institute of Standards and Technology.
• Iqbal, N., & Salley, M. H. (2004). Fire Dynamics Tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program (NUREG-1805). U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation.
• Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
• National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
• Razza, J. C., & Grill, R. A. (2005b). Comparison of codes, standards, and practices in use at the time of the design and construction of World Trade Center 1, 2, and 7 (NIST NCSTAR 1-1E). National Institute of Standards and Technology.
• Razza, J. C., & Grill, R. A. (2005c). Comparison of the 1968 and current (2003) New York City Building Code provisions (NIST NCSTAR 1-1F). National Institute of Standards and Technology.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.